How Kirimana compares
Honest comparisons. We’re not trying to replace every vendor in your stack — we sit above them and operationalise the contract layer. Below is where each vendor wins, where they don’t, and what Kirimana adds that none of them ship.
What makes Kirimana unique
The open-source, AI-native data contract and automation platform.
- Open source (Apache-2.0). Atlan, Collibra, Gable, Datatera are not. Unity Catalog, Purview, Horizon are vendor-native.
- AI-native. Every contract has an AI policy. Every LLM call is gated by classification. Not bolted on after the GA — built in from contract zero.
- Multi-platform. The same ODCS contract runs on Databricks, Fabric, Trino, DuckDB, Postgres. The platform adapter is ~400 lines.
- Contract-driven, not catalog-driven. Atlan, Collibra, Alation, Unity Catalog, Purview observe what exists. Kirimana decides what should exist before anything is built.
- End-to-end. From contract draft → apply → audit → compliance report. Gable / Datatera / Schemata stop at validation; Kirimana goes the full distance.
- Ships compliance generators. DORA, EU AI Act, GDPR Art. 17 redaction reports built in. None of the adjacent vendors ship these out of the box.
- Catalog pass-through, not catalog replacement. We push to Unity Catalog / Purview / Horizon / Polaris. The catalog stays yours; the contract is ours.
- Six platform-agnostic USPs that compound at scale: per-contract AI policy, contract state machine, goal-to-data lineage, federated library, PR-time linting, non-technical governance UI.
Detailed comparison
vs Atlan / Collibra / Alation (data catalogs)
| Kirimana | Atlan | Collibra | Alation | |
|---|---|---|---|---|
| Category | Data contract platform | Data catalog | Data governance + catalog | Data catalog |
| Open source | ✓ Apache-2.0 | ✗ | ✗ | ✗ |
| AI-native | ✓ built in | bolted on | bolted on | bolted on |
| Multi-platform | ✓ | ✓ | ✓ | ✓ |
| Contract artefact | ✓ ODCS canonical | ✗ | ✗ | ✗ |
| AI policy enforcement per contract | ✓ classification-gated | manual | manual | manual |
| Contract state machine | ✓ | ✗ | ✗ | ✗ |
| PR-time linting | ✓ | ✗ | ✗ | ✗ |
| DORA + EU AI Act + GDPR generators | ✓ built-in | add-on | add-on | partial |
| Catalog UX polish | basic | ✓ excellent | enterprise | enterprise |
| Pricing | Free | $$$ | $$$$ | $$$ |
How we differ: Atlan wins on catalog UX. We don’t compete there — we feed Atlan via push adapter (Pro Services) so you can have both. Collibra owns deep enterprise governance suites; we’re lighter, AI-native, and free.
vs dbt Cloud (transformation orchestration)
| Kirimana | dbt Cloud | |
|---|---|---|
| Category | Data contract platform | Transformation orchestration |
| Open source | ✓ Apache-2.0 | dbt-core OSS, Cloud is paid |
| Engine | wraps dbt-core | own dbt runtime |
| Contract artefact | ✓ ODCS canonical | model.yml (lightweight) |
| AI policy enforcement per contract | ✓ | ✗ |
| Multi-platform adapters | ✓ all major | ✓ via dbt adapters |
| Goal-to-data lineage | ✓ | model lineage only |
| DORA + EU AI Act + GDPR generators | ✓ | ✗ |
How we differ: We don’t compete with dbt-core. We wrap
dbt-core; you keep running the same dbt build. We add the
contract, AI policy, audit, and compliance layer above it.
vs Gable.ai / Datatera / Schemata (data-contract specialists)
| Kirimana | Gable.ai | Datatera | Schemata | |
|---|---|---|---|---|
| Category | Data contract platform | Contract validator | Contract validator | Contract validator |
| Open source | ✓ Apache-2.0 | ✗ | ✗ | ✗ |
| End-to-end (apply, audit, dispatch) | ✓ | partial | partial | ✗ validation only |
| Multi-platform adapters | ✓ all major | limited | limited | limited |
| AI-native | ✓ | partial | partial | ✗ |
| PR-time linting | ✓ | ✓ | ✓ | ✓ |
| Catalog pass-through | ✓ Unity / Purview / Horizon / Polaris | partial | partial | partial |
| Compliance generators | ✓ DORA / EU AI Act / GDPR | ✗ | ✗ | ✗ |
| Federated library | ✓ | ✗ | ✗ | ✗ |
| MCP for AI assistants | ✓ | ✗ | ✗ | ✗ |
How we differ: the data-contract specialists validate. We operationalize. Gable raised on the contract-validation thesis; Kirimana is a superset that includes validation plus apply, dispatch, audit, redaction, AI gating, and compliance reporting.
vs Unity Catalog / Microsoft Purview / Snowflake Horizon (vendor-native governance)
| Kirimana | Unity Catalog | Microsoft Purview | Snowflake Horizon | |
|---|---|---|---|---|
| Category | Data contract platform | Vendor-native catalog | Vendor-native catalog | Vendor-native governance |
| Open source | ✓ | ✗ | ✗ | ✗ |
| Cloud-agnostic | ✓ | Databricks-only | Microsoft-only | Snowflake-only |
| Contract artefact | ✓ ODCS canonical | tags + lineage | tags + lineage | tags + lineage |
| AI policy enforcement per contract | ✓ | ✗ | ✗ | ✗ |
| Contract state machine | ✓ | ✗ | ✗ | ✗ |
| DORA + EU AI Act + GDPR generators | ✓ built-in | ✗ | partial (Microsoft Compliance Manager) | ✗ |
| Compatible with each other | ✓ feeds them | ✗ | ✗ | ✗ |
How we differ: We integrate, we don’t replace. Unity Catalog stays for Databricks shops; Purview stays for Microsoft; Horizon stays for Snowflake. Kirimana feeds them via push + optional pull, so you keep their UX and add the contract + governance + AI-policy layer above.
vs Monte Carlo / Soda / Bigeye (data observability)
| Kirimana | Monte Carlo | Soda | Bigeye | |
|---|---|---|---|---|
| Category | Contract platform | Observability | Observability + tests | Observability |
| Open source | ✓ | ✗ | partial | ✗ |
| Reactive detection | partial | ✓ | ✓ | ✓ |
| Contract-driven monitoring | ✓ | ✗ | partial | ✗ |
| AI policy enforcement | ✓ | ✗ | ✗ | ✗ |
How we differ: observability vendors detect when something is wrong. Kirimana prevents what shouldn’t be allowed (via the contract + PR-time linter + AI gate) and records what has to happen (via the audit log + redaction surface). They’re adjacent; you can run both.
Where Kirimana might not be right for you
- You want catalog UX polish first. Atlan still wins on UX. Run Kirimana + Atlan together (Pro Services adapter shelf).
- You’re a 1-engineer prototype. Use dbt-core directly. Add Kirimana when contracts start mattering — usually around team-size 5+.
- You only need transformation orchestration. dbt-core (free) or dbt Cloud is enough. Don’t pay for a contract layer you won’t use.
- You only need observability. Run Soda or Monte Carlo alongside Kirimana, not instead.
Where Kirimana is the right call
- Mid-size to enterprise data teams that need contracts to matter, not just be paperwork.
- Regulated industries where DORA / EU AI Act / GDPR reporting can’t be reverse-engineered from the catalog.
- Multi-platform estates (Databricks + Fabric + Trino) where one governance story has to travel.
- AI-heavy teams where every LLM call needs a policy and an audit row.
- OSS-first organisations that need to fork, audit, self-host.
- Public sector + sovereignty-critical buyers that can’t put governance on a US-vendor SaaS.
Talk to Kiri
Kiri can answer specific comparison questions — “how does Kirimana’s AI policy gate compare to Unity Catalog’s attribute-based access control?”, “would Atlan + Kirimana give me both the UX and the contracts?”. Ask: /chat.